Effective Migration of an Automation System to Microservice Architecture

The complexity of software systems has increased, which has led to increased interest toward microservices as a style of software architecture, where the application is broken down into small independent services, built around business capabilities. The migration process from a monolithic architecture to a microservice architecture is a complex one and includes multiple influencing factors. As microservices are also gaining attention outside software industries there are business environment specific influencing factors that have not been previously recognized. The goal of this thesis is to study how the migration process from a monolithic architecture to a microservice architecture can be done effectively. This is done by mapping out the different influencing factors in the migration process and by providing guidelines for the different phases in the migration process. This study approaches the microservices migration process from a new angle, focusing on the different aspects (technological, organizational, environmental) of the process. The study recog-nizes the impact that the business environment can have on the migration process. This is im-portant as microservices are being adopted in different industries and the unique aspects of those industries can affect the feasibility of microservices. The study was conducted as a single case study for a large Finnish industrial organization that is planning to migrate their existing system to microservices. In this study 14 employees with varying roles from the case organization were interviewed. The focus of the interviews was on collecting information about the main influencing factors the interviewees felt would affect the migration process. The interviews were also used to collect information about the challenges in the current system, that serve as focus points for challenges that microservices aim to partly solve. There exist multiple influencing factors that can be categorized to technological, organiza-tional, and environmental groups. The environmental context has not been previously studied. The results indicate that the environmental context can limit the feasibility of microservices by setting limitations for the technological and organizational solutions. The results also indicate that the migration process involves multiple stakeholders with varying interests, who may not under-stand each other’s point-of-views. Contrast was found especially between the benefits perceived from the point-of-view of software development and the long-term business benefits. To effec-tively migrate to microservices, commitment from every stakeholder is required, which can be achieved by working jointly toward a common goal that creates benefits for everyone